Project 1 Abstract Black women suffer higher incidence of poor-prognosis breast cancer subtypes and worse stage-specific mortality. Our previous Carolina Breast Cancer Study(CBCS)-based SPORE discoveries have integrated population-based epidemiology with tumor biology to demonstrate that young black women are more likely to develop Basal-like tumors. More recently, we found that black women also have higher frequency of Luminal B and HER2-enriched breast tumors. Furthermore, even among the best-prognosis clinical subset of estrogen receptor (ER)-positive/HER2-negative cancers, black women have higher risk of recurrence (ROR) scores. These findings implicate tumor biology as a key contributor to racial mortality disparities. Some tumor factors (i.e., higher frequency of HER2-enriched subtype) may be targetable, illustrating how population-based genomics can advance precision medicine. In this Project, we propose to deepen our understanding of racial differences in tumor biology, etiology and progression, with a focus on tumor mutational signatures and immune responses. The frequency of specific mutations (i.e. TP53, PIK3CA), and the overall patterns of somatic mutations (mutational signatures) can be informative about the underlying biological processes that have gone awry from the earliest stages of carcinogenesis. In addition, new discoveries have emphasized the importance of the immune system in breast cancer etiology and progression. With our rich resource of tumor biospecimens (exceptional 96% tumor procurement from 3000 cases), we will sequence tumor DNA and identify mutational signatures in black and white women (Aim 1) and we will characterize tumor immune microenvironments using six immunohistochemical markers and RNA profiling for 50 genes (Aim 2). While The Cancer Genome Atlas (TCGA) has evaluated racial differences in tumors with hundreds of thousands of data points per patient, limitations included small numbers of black participants (CBCS is ten-fold larger than TCGA), a sampling scheme that was not population-based, and lack of detailed covariate and follow-up data. TCGA advances have allowed us to increase sequencing efficiency for population-based work and have helped in developing an efficient RNA panel for immune profiling, resulting in a plan for cost-effective collection of mutational signatures and expression profiles in 2000 women (1000 black, 1000 white, sampling from the entirety of CBCS). Utilizing these advances in the context of the CBCS, we hypothesize that mutational signatures and immune responses are differentially associated with race, age, and breast cancer subtype. We also expect to identify some mutational signatures that vary as a function of immune profiles. The tumor biological variables collected in Aims 1 and 2 will be assessed as predictors of breast cancer recurrence and survival. In addition, impact of tumor biology may be confounded by other patient-level factors such as socioeconomic status, access to care, treatment delay and treatment adherence. Multivariable modeling (Aim 3) will identify the relative strengths of tumor biology and patient-level features that contribute to survival, thus allowing for impactful precision medicine strategies. 1